Radioimmunotherapy (RAIT) using a humanized murine monoclonal antibody, chimeric L6 (ChL6), has produced objective tumor reduction in 50% of chemotherapy-refractory patients with metastatic breast cancer in our prior studies. Because myelosuppression limited dose escalation, we evaluated the ability of granulocyte colony-stimulating factor (G-CSF)-mobilized peripheral blood progenitor cell (PBPC) transfusions to ameliorate this problem. 131I-labeled ChL6 was given at a starting dose of 150 mCi/m2 (2,5 times the maximum tolerated dose without PBPCs) for a planned three treatments. When blood radioactivity declined to less than 1 microCi/ml after treatment, PBPCs were transfused, and G-CSF was administered. Patient 1 had minimal myelosuppression, received two cycles of therapy, and then developed human antimonoclonal antibody (HAMA). Patient 2 had prolonged thrombocytopenia that resolved after additional PBPC transfusion. Progressive disease as well as HAMA prevented further treatment. Patient 3 received all three cycles of 150 mCi/m2 at 8-week intervals. Thrombocytopenia (< 25,000/microliter) occurred but was transient (0-7 days). Because HAMA developed in all prior patients who received G-CSF with ChL6 RAIT, including patients 1 and 2, who received PBPC, patient 3 was given cyclosporin for 14 days. She did not develop HAMA or significant toxicity following 3 cycles of RAIT. Cumulative radiation doses to her lungs and tumor were estimated at 3,100 and 11,200 cGy, respectively. For 9 months, she had a reduction in bone pain, a decline in serum tumor markers, and decreased tumor uptake of F-18-deoxyglucose on a positron emission scan. Her performance status improved, and she had no pulmonary toxicity. We conclude that: (a) PBPC transfusion can modify the myelotoxicity of RAIT and can permit repetitive dosing; (b) cyclosporin is a promising means to abrogate HAMA; and (c) fractionation of intensive-dose RAIT may increase the antitumor effect and reduce normal organ toxicity.